trainer_state.json

{
  "best_metric": 0.9314666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0236/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 9.602340698242188,
      "learning_rate": 4.7538820348083834e-05,
      "loss": 0.779,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.4564119875431061,
      "eval_runtime": 7.4615,
      "eval_samples_per_second": 502.581,
      "eval_steps_per_second": 7.907,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.362864017486572,
      "learning_rate": 4.063987381465584e-05,
      "loss": 0.2418,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8877333333333334,
      "eval_loss": 0.36500826478004456,
      "eval_runtime": 7.7935,
      "eval_samples_per_second": 481.168,
      "eval_steps_per_second": 7.57,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 131.2980194091797,
      "learning_rate": 3.0628700751103485e-05,
      "loss": 0.1206,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.37638723850250244,
      "eval_runtime": 7.368,
      "eval_samples_per_second": 508.961,
      "eval_steps_per_second": 8.008,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 37.084922790527344,
      "learning_rate": 1.950269447231056e-05,
      "loss": 0.0563,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9066666666666666,
      "eval_loss": 0.3469636142253876,
      "eval_runtime": 7.8638,
      "eval_samples_per_second": 476.867,
      "eval_steps_per_second": 7.503,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 9.491910332815573e-06,
      "loss": 0.0246,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9232,
      "eval_loss": 0.28639206290245056,
      "eval_runtime": 7.6456,
      "eval_samples_per_second": 490.479,
      "eval_steps_per_second": 7.717,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.06184566766023636,
      "learning_rate": 2.5198196276040782e-06,
      "loss": 0.0056,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9293333333333333,
      "eval_loss": 0.2943213880062103,
      "eval_runtime": 7.5918,
      "eval_samples_per_second": 493.953,
      "eval_steps_per_second": 7.772,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 7.793915748596191,
      "learning_rate": 2.0434646496314548e-10,
      "loss": 0.0018,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9314666666666667,
      "eval_loss": 0.2768959701061249,
      "eval_runtime": 7.41,
      "eval_samples_per_second": 506.075,
      "eval_steps_per_second": 7.962,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}